DE68913550T2 - GEOPHONE ARRANGEMENT. - Google Patents

Description

The invention relates to a geophone system for seismic exploration with a plurality of geophones, each of which consists of a mechanical transducer and an electronic processing circuit and is connected to a central processing station by means of a transmission line

To measure seismic signals in such a system, seismometers or geophones are used. These geophones are generally passive analog sensors that are connected in series in groups and to a control room. When the geophone moves, a voltage is induced in a coil that is suspended in the magnetic field of a permanent magnet. To ensure high sensitivity, the mass of the magnet is large, which has an adverse effect on the coupling at high frequencies between the geophone and the ground in which it is inserted. The analog connection between the geophones and the control room is sensitive to interference from external electromagnetic fields.

In the processing station, the analog output signal of the geophones is amplified, sampled and digitized. Given the high demands on resolution, the analog-digital converter and the noise filter required for this purpose are so sensitive to component tolerances that manufacturing them using IC technology is almost impossible.

Geophones that are not working properly can have a significant impact on the seismic data collected - an impact that only becomes apparent during data processing, at which stage the recovery of said data is no longer possible. To prevent this, geophones are tested in field use.

From EP-A O 264 509 (Schlumberger) a system for the exploration of boreholes of the type specified in the preamble of claim 1 is known, with measuring transducers, in in which the position of an inertial mass is detected by optical means which, via an amplifier, control the current flowing through the coil of the transducer in question in order to return the said mass to its original position.

However, the operation of such a transducer depends on its orientation in the gravitational field, which also requires a wide measuring range. For stability reasons, the loop gain of the processing circuit must be low. Such a system is not very suitable for practical seismic exploration, but the low loop gain makes it impossible to digitize such a system, since in this case a high loop gain is required to suppress the quantization noise. Digitization is very advantageous in terms of the suppression of signal interference in the line or lines leading to the central data processing station, and also has other advantages in terms of data processing and the grouping of the various sensors in the overall system.

It is an object of the invention to provide a system of this type free from these disadvantages, the features of which are mentioned in the characterising part of claim 1. This system has the following properties: due to the digital data transmission, the influence of interference on the cable will be small and the distance between the geophone and the control room can be large; the analogue-digital converter and the interference filter are so insensitive to component tolerances that their manufacture by IC technology is possible and the analogue-digital converter and the interference filter can therefore be built into the geophone; the bandwidth of the geophone is large and is determined by a digital filter which ensures a high degree of freedom; the inertial mass is small and the geophone therefore has a small mass and a small volume, which allows good Ground coupling is ensured; given the small deviation of the inertial mass, springs with high transverse stiffness can be used; the geophone can be used in any position and only responds to axial vibration; the transmission and distortion of the geophone can be measured using a digital test signal.

From "A High Resolution Wideband Digital Feedback System for Seismometers" J. Phys. Sci. Instr. 21 (1988,08) 748-752, an article by Z. Yin and M.J. Usher, an acceleration transducer with a digital output is known. This transducer is equipped with a capacitive sensor for detecting the displacement of an inertial mass, an analog-to-digital converter for generating a digital signal and a magnetic coil force feedback transducer controlled by a digital-to-analog converter. By changing the parameters in a computer program it is possible to set the gain range and vary the bandwidth. To prevent instability of the transducer system, pole zero cancellation is used. Pole zero cancellation requires high component accuracy and is not suitable for a system with high loop gain and wide bandwidth, which is to be manufactured using IC technology. The reduction of quantization noise in this system is therefore relatively low. In addition, this known system is not intended for seismic exploration in practical use with the help of a plurality of distributed sensors, but for seismological investigations relating to earthquakes and similar geological phenomena and comprises only one sensor in a laboratory setup which is directly connected to a data processing device.

From US-A 3 429 189 (HF Krabbe) an accelerometer is known that provides a digital output signal. The sensor group consists of a Sensing element which determines the position of an inertial mass, and a control coil which exerts a restoring force on the inertial mass. When the output of the position sensing element exceeds a positive or negative value, a current pulse is transmitted through the control coil. This results in a Lorentz force being exerted on the inertial mass which is opposite to the force caused by the acceleration to be measured. The motion of the inertial mass is reduced to essentially zero by the restoring force. Since the duration of the current pulse is short and constant, the sum of the current pulses is proportional to the average Lorentz force and it is therefore possible to calculate the acceleration. The speed of the input signal to be measured is proportional to the frequency of the output signal. In effect, a sensor with a frequency output is obtained in this way.

This accelerometer is not suitable for seismic exploration. To achieve a high dynamic range, the loop gain should be set very high, but this would result in significantly lower attenuation of the sensor array and instability.

This known accelerometer can also only be tested using a mechanical input signal, which is undesirable given the large number of geophones used. A geophone is known from NL-A 82 02 309 in which a coil is used as a sensing element to detect the movement of an inertial mass, while a second coil controls a force transducer. The mutual inductance of the two coils limits the loop gain. The mechanical design is complicated because the magnets have to be magnetized before they are glued to an H-shaped pole piece, which is only possible in a clean room. In addition, it is difficult to meet the mechanical tolerances of the H-shaped pole piece. This Sensor does not provide a digital output signal and does not use a digital signal as a test signal.

US-A 3 088 062 (Hudimac) describes a measuring device for a sonar system in which the movement of an inertial mass is damped by the Lorentz force generated by a current flowing through two coils connected in series, said coils being connected to a negative impedance so that the current flowing in it corresponds to the movement. Since the fluctuations in the magnetic flux in the coils are caused by variations in the air gap, the linearity is too limited for seismic exploration.

The invention is described in more detail below with reference to the drawing. It shows

Figure 1 shows a circuit diagram of the digital geophone according to the invention;

Figure 2 shows a circuit diagram of the digital geophone, in which improved noise suppression is achieved by means of a second feedback loop;

Figure 3 shows a circuit diagram of the digital geophone, where a higher sampling frequency is possible;

Figure 4 shows the frequency characteristics of the input signal and the test signal, also showing the influence of the quantization noise;

Figure 5 shows the mechanical transducer provided in the geophone;

Figure 6 a mechanical transducer that can be used for the geophone and in which the centering forces acting on the inertial mass are absorbed by a rigid construction; and

Figure 7 shows a circuit element of the circuit with which the geophone can be tested; this circuit measures the harmonic distortion for this purpose.

Figure 1 shows a circuit diagram of a digital geophone I according to the invention. The mechanical signals are shown by dashed lines and the electrical signals by solid lines.

This geophone 1 comprises an inertial mass 2 forming part of a mechanical transducer 3, said mass being supported by damped springs shown in simplified diagrammatic form.

When the geophone is under the influence of an acceleration input signal X(s), a force Fa acts on the mass 2. If the resultant force of this force Fa and the Lorentz force F1 explained below is not zero, this results in a movement of the inertial mass 2 which is detected by the speed sensor 4. The operation of the speed sensor 4 is explained in more detail below with reference to Figures 5 and 6. In physical terms, the sensing element 4 has an output voltage, but since it is designed as an integrated circuit, this voltage is converted into an output current. The output signal of the sensing element 4 is amplified by an amplifier 5. The signal is sampled by the sampling element 6 after the "HOLD" command is given from a clock generator 7. The clock generator 7 is controlled by an external synchronization signal "SYNC". The sampled signal is converted into a digital signal by the analog-digital converter 8 after the "START" command has been issued by the clock generator 7. The sampling frequency fs is high and in any case much higher than the resonance frequency of the measuring transducer 3. Due to the limited resolution of the analog-digital converter 8, it can be designed as an integrated circuit. The digital output signal Y(z) is inverted by an inverter 9 of a feedback circuit, so that the signal in the feedback circuit is in phase opposition to the input signal. A digital sweep stage 10 adds a test signal T(z) to the inverted output signal. During measurement, the signal T(z) is zero. The sum signal of the summing stage 10 is converted by the digital-analog converter 11 into a current i2 fed to a force measuring transducer 12, which then applies a current to the current to the inertial mass 2. i2 exerts a Lorentz force F1 proportional to the mass 2. As a result of the Lorentz force, the motion of the inertial mass 2 is essentially reduced to zero.

In order for geophones to be suitable for seismic measurements, it must be possible to test the geophone. Testing of sensor group 1 is explained in more detail below with reference to Figure 7. The transfer function of the digital geophone can be tested using the test signal T(z). The test signal is summed with the inverted output signal using the digital summing stage 10 and transmitted by the digital-to-analog converter 11 through the force transducer 12. The inertial mass 2 is excited in the same way as by the input acceleration signal. The transfer function Y(z)/T(z) is, apart from a frequency-independent factor, almost equal to the transfer function Y(s)/X(s) in the seismic bandwidth. In order to use the test device, it is necessary that the output signal and the test signal are synchronized. Therefore, the clock generator generates the "START" command using the external synchronization signal.

In order to reduce the number of wires, a common data line can be used between a group of geophones and a control room. Serial data transmission is used to transmit the measurement data to the control room, and a control signal can be issued to control the geophones. This control signal includes the address of the geophone for which it is intended, as well as the test signal and the corresponding synchronization signal in digital form. A geophone transmits its measurement data to the control room after recognizing its address. Such a data transmission system is known and will therefore not be explained in more detail.

Figure 2 shows an alternative circuit diagram of the digital geophone. The output current of the speed sensor element 4 is now summed with a current i1 discussed below and divided by the Amplifier 5'. The amplifier 5 has a very high gain factor and a low-pass transfer characteristic. The limit point of the frequency characteristic is very low, so that the input signal is integrated. The output signal of the low-pass amplifier 5' is sampled by the sampling element 6 after the "HOLD" command is issued. The analog-digital converter 8 converts the voltage into a digital output signal Y(z), and after inversion by the inverter 9, the output signal Y(z) is added to the test signal T(z). The resulting sum signal is converted by the digital-analog converter 11 into two output currents i1 and i2. The current il is added to the output current of the speed sensor element 4 and integrated by the low-pass amplifier 5'. The force transducer 12 exerts a Lorentz force F1 proportional to the current i2 on the inertial mass 2.

The gain factors of the low-pass amplifier 5' should be high. The feedback in the sensor group 1 for a signal with a frequency of zero is interrupted because the speed sensor 4 can only detect a movement of the inertial mass 2. The offset voltage of the low-pass amplifier 5' appears with a very high gain factor at the output and thus limits the dynamic range of the sensor. In IC technology, offset compensation is difficult to achieve and is therefore complex. This disadvantage is limited by arranging a second feedback circuit with the current i1. The offset of the low-pass amplifier 5' is now added to the inverted output signal, whereby the output signal is effectively reduced as a result of the offset. The gain factor of the low-pass amplifier 5' can now be selected very high without the offset limiting the dynamic range.

In addition, the second feedback circuit provides better suppression of quantization noise. This is explained in more detail below.

The digital-analog converter 11 with two output currents can be provided with the help of a current mirror with multiple output according to IC technology.

Figure 3 shows an alternative circuit diagram. According to this diagram, the currents i1 and i2 are generated by means of two separate digital-analog converters 11 and 11'. Using the device according to this circuit diagram, the sampling frequency used to convert the digital word into a current i1 can be chosen to be much higher than the sampling frequency that generates the current i2. This has the advantage that the sampling frequency is not limited by the maximum frequency of the current i2, in which case the voltage within the control coil 12 is lower than the supply voltage. By choosing a very high sampling frequency for the analog-digital converter 8 and the digital-analog converter 11', components with a lower resolution can be used for this purpose. An arithmetic unit 10' now adds the inverted output signal Y(z) to the test signal T(z), integrates the sum signal and, after a clock pulse from circuit 7, transmits the signal to the digital-analog converter 11. The integration of the sum signal caused by the low-pass characteristic of the arithmetic unit 10' improves the resolution of the coupled signal.

Figure 4 shows the frequency characteristics of the sensor group. The mass-spring system of the transducer 3 has a resonance frequency within the spectrum of the seismic bandwidth, but due to the high open loop gain, the transmission of the sensor group 1 within the seismic bandwidth is hardly influenced by the frequency characteristics of the mechanical transducer 3. At frequencies exceeding the seismic bandwidth, the open loop gain is low due to the low transfer function of the mass-spring system 3 and the low-pass filter 5'. The The transfer function Y(s)/X(s) of sensor group 1 then decreases with increasing frequency. Since the speed sensor element 4 only responds to the movement of the inertial mass 2, the transfer function of sensor group 1 will be low at frequencies below the seismic bandwidth.

The resolution of the analog-to-digital converter 8 is poor and quantization noise is added to the output signal. This quantization noise is evenly distributed over the frequency range, but due to the high open loop gain at low frequencies, the contribution of the quantization noise to the output signal will be small. At higher frequencies, the open loop gain decreases and the quantization noise is no longer suppressed. In Figure 4, the contribution of the quantization noise in the output signal is shown by a hatched area. From this figure it can be seen that the contribution of the quantization noise is strongly suppressed within the seismic bandwidth (up to frequency fc). By choosing a higher sampling frequency or by using a multiple feedback loop, maximum suppression of the quantization noise is achieved.

By means of a digital low-pass filter, frequencies above the frequency fc can be suppressed. As a result of the use of the digital filter, the signal-to-noise ratio increases and the resolution of the digital output signal is improved. It is possible to combine the digital filter of several geophones belonging to a group by adding the geophone data. If the digital signal is sampled again after the low-pass filter, the capacity of the data transmission channel and the data storage is used more effectively. If the filter is made using VLSI technology, it can be installed inside the geophone. This technology is well known in itself and will therefore not be discussed in detail.

Figure 5 shows an embodiment of the mechanical transducer 3, which is designated here in its entirety by the reference number 13. This transducer 13 consists of an inertial mass 2 which comprises a magnet 14, a spacer 15 and inner pole pieces 16, and of a housing 17 with a sensing coil 18, a control coil 19, a compensating coil 20 and the tubular pole piece 21. The inertial mass 2 is suspended in the housing 17 by means of springs 22. A movement of the inertial mass 2 changes the magnetic flux within the sensing coil 18 and subsequently induces a voltage. The sensing coil 18 is connected in series with the compensating coil 20 and connected to the input of the electronic processing device. Since the magnetic field generated by the magnet 14 is zero in the air gap of the compensating coil 21, a movement of the inertial mass 2 does not induce a voltage in the compensating coil 214. The control coil 19 is connected to the electronic processing device and is located in the magnetic field of the magnet 14. If a current is passed through the control coil 19, the Lorentz force F1 = Bil exerts a force on the inertial mass 2. Since the magnetic flux density B and the coil length l are constant, the Lorentz force F1 is proportional to the current i. The current flowing through the control coil 19 also causes an induced voltage to be induced by mutual induction between the control coil 19 and the sensing coil 18. In order to prevent the electronic processing device from responding to this induced voltage as a movement of the inertial mass 2, the compensating coil 20 is provided. Since the relative permeability value of the spacer 15 is equal to that of the magnet 14, ie that of the air, the voltage induced in the compensating coil 20 will be equal to the induced voltage in the sensing coil 18. However, the two coils are in opposite directions wound so that, due to the series connection of the two coils, the induced voltage caused by a current flowing through the control coil 19 is zero.

Capacitive coupling could occur between these coils. To prevent this, the inner pole pieces 16 are connected to each other and to the electrical ground of the electronic processing device by means of a core 23. The material from which the core 23 is made is electrically conductive but not magnetically conductive and consists, for example, of copper. The core 23 also serves to mechanically attach the pole pieces 16, the magnet 14 and the spacer 15 to the springs 22. A mechanical connection by gluing is conceivable, but not recommended in view of the shock resistance.

The mechanical structure according to Figure 5 is simple, but has the disadvantage that the outer pole shoe 21, which consists of a material that conducts the magnetic field, is attracted by the inner pole shoes 15 and the magnet 14. The springs should therefore be sufficiently stiff in the radial direction to develop a centering force sufficient to compensate for the attraction force of the magnet 14. Figure 6 shows another embodiment of the mechanical transducer 13' that does not have this disadvantage. In this case, the inertial mass 2' consists of the magnet 14, a spacer 15, inner pole shoes 16 and an outer pole shoe 21'. The housing 17 comprises the sensing coil 18, the control coil 19 and the compensating coil 20. The outer pole piece 21' is rigidly attached to the outside of the inner pole pieces 16 by means of pole teeth 23'. The magnetic attraction force between the inner pole pieces 16, the magnet 14 and the outer pole piece 21' is now absorbed by the rigid mechanical connection, and the inertial mass 2' is suspended in the housing 17 by means of the springs 22.

Figure 7 illustrates a method for measuring the distortion level of sensor group 1. A read-only memory 24 contains a digitized sinusoidal signal. A switch 25 samples the read-only memory and generates the test signal T(z). The period of the test signal T(z) is determined by the clock frequency generated by the clock generator 26 and an arithmetic unit 27. The frequency of the test signal can be varied by controlling n₁. The test signal T(z?) excites the inertial mass 2 and simulates an acceleration input signal X(s). The output signal Y(z) contains the response of the test signal T(z) and the acceleration signal X(s).

At the same time, a second switch 28 scans the read-only memory 24, the period of the test signal being determined by the clock 26 and an arithmetic unit 29. This test signal is generated with a frequency that is n₂/n₁ times higher than the frequency of the first test signal. This signal is multiplied by a multiplier 30 with the output signal of the sensor 1. The product formed by the multiplier 30 comprises sum and difference frequencies of the output signal Y(z) and the test signal, whose frequency is n₂/n₁ times higher. The harmonic distortion in the output signal, with a frequency n₂/n₁ times higher, can be determined by measuring the difference with respect to a zero frequency at the output of the multiplier 30. The output signal V(z) is the harmonic distortion with a frequency n₂/n₁ times the frequency of the test signal T(z). A low-pass filter 31 filters this frequency out of the signal of the multiplier 30. The cutoff frequency of the low-pass filter 31 determines the width of the filter in the spectrum of the seismic signal. As the cutoff frequency is chosen lower, the contribution of the environmental noise due to the input acceleration decreases. Since the word length of the signals Y(z) and T(z) is small, the fabrication of the digital circuit will be simple.

Claims (9)

1. A geophone system comprising a plurality of geophones (1), each of which consists of a mechanical transducer (3) and an electronic processing circuit (5...11), the mechanical transducer (3) comprising an inertial mass (2) and a force transducer (12), the inertial mass (2) being adapted to be excited by an input acceleration signal and by the force transducer (12), the excitation being detected by a sensing element (4), said force transducer (12) being controlled by the electronic processing circuit (5...11), and the processing circuits of said geophones (1) being connected to a central station by means of a transmission line, characterized in that the sensing element (4) detects the speed of the inertial mass (2) and drives an analog-to-digital converter (8) by means of an amplifier (5, 5') with a very high gain factor, that the output of the analog-digital converter (8) is connected to the input of a digital-analog converter (11), and that an output of this digital-analog converter (11) controls the force transducer (12), whereby the input of the analog-digital converter (8) is sampled at a much higher ratio than the mechanical oscillation frequency of the mechanical transducer (3). 2. The system according to claim 1, characterized in that the amplifier (5') has a low-pass characteristic, and in that its input, in addition to its connection to the speed sensor element (4), is connected to the output of the digital-to-analog converter (11) in order to limit the control of the analog-to-digital converter (8) due to the offset of the amplifier (5') and to suppress the quantization noise. 3. The system according to claim 2, characterized in that the input of the amplifier 5' in addition to its Connection with the output of the sensor element (4) is connected to the output of a second digital-analog converter (11'), that the input of the second digital-analog converter (11') is controlled by an inverter (9), and that the inverter (9) simultaneously controls an arithmetic unit (10') which controls the first digital-analog converter (11) after a sampling command from a clock generator (7), wherein the arithmetic unit (10') integrates the digital signal. 4. The system according to claim 1, 2 or 3, characterized in that a digital test signal T(z) is added to the output signal of the analog-digital converter (8) in a summing stage (10, 10'), the output of this summing stage (10, 10') being connected to the digital-analog converter (11) in order to allow a force to act on the inertial mass (2) in accordance with this test signal, and by means of the force transducer (12) the clock signal (7) for the analog-digital converter simultaneously acts as a clock signal for the test signal. 5. The system according to any one of claims 1...4, in which each mechanical transducer (3) comprises a coaxial permanent magnet group (14, 21) and a coaxial coil group comprising a sensing coil (18) and a control coil (19) coaxially arranged in a tubular outer pole piece (21), one assembly being movable and the other stationary, and springs (22) being provided between both assemblies, characterized in that the magnet group consists of a permanent magnet (14) provided on both sides with two inner pole pieces (16), the control coil (19) is arranged in the air gap between one of the inner pole pieces (16) and the outer pole piece (21) and the sensing coil (18) is arranged in the air gap between the other inner pole piece (16) and the outer pole piece (21), and a compensating coil (20) in series with the sensing coil (18) outside the magnetic field of the magnet (14) and in the outer pole piece (21), all in such Such that the induction voltage induced in the compensating coil (20) as a result of the alternating flux generated by the control coil (19) is substantially equal to the induction voltage induced in the sensing coil (18) as a result of the alternating flux. 6. The system according to claim 5, characterized in that the outer pole shoe (21') is mechanically rigidly connected to the inner pole shoe (16) by means of two pole teeth (23'). 7 The system according to claim 5 or 6, characterized in that the magnet group comprises a third inner pole shoe (16) in front of the compensating coil (20) and is separated from the first inner pole shoe (16) by a non-magnetic gap (15). 8. The system according to claim 5, 6 or 7, characterized in that the inner pole shoes (16) are connected to one another by means of an electrically conductive core (23) directed in the axial direction of the assembly. 9. An apparatus for testing the geophone system of claims 1, 2 or 3, characterized in that a digital test signal (2) is generated by reading a signal stored in a memory for controlling the geophone, that a second test signal U(z) is generated by reading said memory at an n-times higher speed, a digital multiplier (30) multiplies said second test signal by the output signal, and the low-pass filter filters out the n-th harmonic distortion from the output signal of the multiplier (30).